Long-TERM goals are to elucidate mechanisms by which human and bacterial mismatch-repair (MMR) systems antagonize ultraviolet-light (UV) mutagenesis, and reduce UV-induced skin cancer. The hypothesis, based on findings (i) that MMR protein specifically recognize mismatched, but not "matched" photoproducts in DNA, and (ii) that UV mutagenesis is increased in MMR-deficient cells, is as follows: During semi-conservative DNA replication, MMR systems recognize incorrect bases incorporated bases incorporated opposite UV photoproducts in template DNA, and direct excision and resynthesis to the nascent error- encoding strands, while preventing nucleotide excision repair (NER) of the mismatched photoproducts. Most studies will employ the highly conserved bacterial and conserved bacterial and human MMR systems in parallel, to elucidate mechanistic differences relevant to human MMR polymorphisms. Effects of their respective MutL homologs on binding affinities of human (MSH) and bacterial (MutS) proteins for mismatched cyclobutane dimers (CPDs) and [6-4] photoproducts in synthetic DNA oligomers will be measured. Using mismatch-repair-proficient human and bacterial cell-free extracts, the relative abilities of the respective mismatched photoproducts and base-mismatch controls to initiate excision tracts in MMR-substrate plasmids, will be compared. The same extracts and substrates will be used to compare the efficiencies and accuracies with which the photoproducts are bypassed during excision- gap-filling DNA resynthesis. The effects of known translesion-synthesis activities-E. coli UmuD'2C protein, human DNA polymerase delta-on bypass efficiency and accuracy will be determined as well. The abilities of MMR proteins to complete with NER proteins for mismatched- photoproduct substrates, thus preventing mutation-fixation by the latter, will also be tested.